In vitro whole blood canine assays were used to quantify the inhibitory actions of the novel non-steroidal anti-inflammatory drug (NSAID) robenacoxib on the cyclooxygenase (COX) isoenzymes, COX-1 and COX-2, in comparison with other drugs of the NSAID class. COX-1 activity was determined by measuring serum thromboxane (Tx)B(2) synthesis in blood samples allowed to clot at 37 degrees C for 1h. COX-2 activity was determined by measuring prostaglandin (PG)E(2) synthesis in blood samples incubated at 37 degrees C for 24h in the presence of lipopolysaccharide. The rank order of selectivity for inhibition of COX-2 versus COX-1 (IC(50) COX-1:IC(50) COX-2) for veterinary drugs was highest with robenacoxib (128.8) compared to deracoxib (48.5), nimesulide (29.2), S+ carprofen (17.6), meloxicam (7.3), etodolac (6.6), R- carprofen (5.8) and ketoprofen (0.88). Selectivity expressed as the clinically relevant ratio IC(20) COX-1:IC(80) COX-2 was highest for robenacoxib (19.8) compared to deracoxib (2.3), S+ carprofen (2.5), R- carprofen (2.1), nimesulide (1.8), etodolac (0.76), meloxicam (0.46) and ketoprofen (0.21). An in vivo pharmacokinetic ex vivo pharmacodynamic study in the dog established dosage and concentration-effect relationships for single oral doses of robenacoxib over the dosage range 0.5-8.0mg/kg. Values of C(max) and AUC were linearly related to dosage over the tested range. Robenacoxib did not inhibit serum TxB(2) synthesis (COX-1) ex vivo at dosages of 0.5-4.0mg/kg and produced only transient inhibition (at the 1h and 2h sampling times) at the 8mg/kg dosage. All dosages of robenacoxib (0.5-8mg/kg) produced marked, significant and dose related inhibition of PGE(2) synthesis (COX-2) ex vivo. The data demonstrate that in the dog robenacoxib is a highly selective inhibitor of the COX-2 isoform of COX, and significantly inhibits COX-2 and spares COX-1 in vivo when administered orally over the dosage range 0.5-4.0mg/kg.